Abstract
Spatial interpolation is employed to improve exposure estimates and to assess adverse health effects associated with environmental risk factors. Since various studies have reported that high ozone (O₃) concentrations can give rise to adverse effects on respiratory symptoms and lung function, we investigated the association between O₃ levels and lung function using a variety of spatial interpolation techniques and evaluated how different methods for estimating exposure may influence health results for a cohort from an industrial complex (Gwangyang Bay) in South Korea in 2009. To estimate daily concentrations of O₃ in each subject, four different methods were used, which include simple averaging, nearest neighbor, inverse distance weighting, and kriging. Also, to compare the association between O₃ levels and lung function by age-groups, we explored ozone's impacts on three age-related groups: children (9-14 years), adults (15-64 years), and the elderly (≥65 years). The overall change of effect size on lung function in each age group tended to show similar patterns for lag and methods for estimating exposure. A significant negative association was only observed between O₃ levels and FVC and FEV₁ for most of the lag and methods in children. The largest effect of O₃ levels was found at the average for the lung function test day and last 2 days (0-2 days). In conclusions, the spatial interpolation methods may benefit in providing individual-level exposure with appropriate temporal resolution from ambient monitors. However, time-activity patterns of residents, monitoring site locations, methodological choices, and other factors should be considered to minimize exposure misclassification.